OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.


Help: OASIS Mailing Lists Help | MarkMail Help

office message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]

Subject: RE: RE: [office] Our Position on the Conformance Proposal

This is my elaboration on the compromise approach that Michael Brauer just sketched
(http://lists.oasis-open.org/archives/office/200902/msg00249.html).  While each of us may attach different importance and rationale to the dual-level approach, I continue to support it.

 - Dennis


We need to remind ourselves that the OASIS ODF 1.0, ISO/IEC 26300:2005, and OASIS 1.1 standards all define what is proposed to be the extended document level as the only standard conformant document level.  So the status quo at this moment is that extensions via foreign elements and attributes and by other specific means are all allowed.

I have no idea why that is the precedent, but it is the precedent and it was subject to all of the scrutiny and deliberations that were involved in advancing those versions of the ODF specification to their current status as (international) standards.  

Whether we are satisfied with that or not, we are faced with a legacy situation that was left to us by the "founding fathers" of ODF.  It falls to us to demonstrate care in the transition to what may be a preferable level of conformance for a broad community, perhaps the broadest community, of ODF users and their chosen software products.  However, this is an open standard in the sense that many point to with great pride.  A consequence is that we have no idea what all the ways the standard has been taken up and applied, all around the globe.  There is no requirement for implementers and those who make solutions in their areas of application to consult or advise us about that.  And even if we were to withdraw earlier standards as pre-empted by 1.2, it would take some time for that to ripple through the community.  I believe that, no matter what our own navels reveal about what is safe to break in going to 1.2, that we should err on the side of caution always and make transitions, not eliminations, and from that assess appropriate future steps.

We have no advance assurance that a universal preference, with procurement teeth, for the proposed pure conformance level will indeed take place, and we need to be careful.  I, for one, am personally satisfied that the dual level for conformant documents will allow us to discover what the larger community of stakeholders will find acceptable, giving us useful feedback for further transition beyond 1.2.  (I still think the namings and the potential for confusion are too great, but prefer to work out the conformance language in the remainder of 1.2 before seeing if there is some reconciliation that can be made with that.  I am not interested in debating that until we are able to see how the details in 1.2 are lined up with the conformance levels.)

It is already possible to specify a procurement and deployment policy requiring limitation to the "strict" schema, although that is  not supported by a defined conformance level in the current specifications and it does not directly forbid foreign elements and attributes (which are extra-schema provisions using either schema), absent a policy.  Note that ODF 1.1 Section 1.5 does not mention the strict schema at all and OFD 1.1 section 1.4 only mentions it with regard to the use of arbitrary content within meta information elements and formatting properties elements.  Section 1.4 is very explicit that the schema that is carried in the body of the specification is "the normative XML Schema for the OpenDocument format."

Now we are moving toward introduction of a normative definition of such a pure (extension-free) level, although it will take more than just the conformance section to accomplish that.  (The remaining extension mechanisms that are separate from those involving foreign elements and attributes are yet to be addressed and identified with a conformance level in the specification.)

It has been said that we can, of course, make ODF 1.2 as different from ODF 1.1 as we choose, and we can change the ceiling on conformance any way we please because that won't change the status of ODF 1.1 and earlier documents.  That is technically true.  It is not a practical truth however, because it ignores the social contract, however weakly it is being handled, with regard to upward compatibility, the assurance that ODF documents will be preservable and processable into the indefinite future, and other expectations that are the context in which the work of the ODF TC will be appraised.  I have no idea how that will play out, but I do think it involves greater issues and concerns, by more stakeholders than those of us engaged in this debate.

-----Original Message-----
From: Warren Turkal [mailto:turkal@google.com] 
Sent: Friday, February 27, 2009 08:47
To: Doug Mahugh
Cc: jomar.silva@br.odfalliance.org; office@lists.oasis-open.org
Subject: Re: RE: [office] Our Position on the Conformance Proposal

I think that there is value in knowing that a document contains no
extensions. In fact, I think producers should aim to use as few
extensions as possible to do what they want. If documents are produced
without extensions, it should be easier for consumers to implement the
functionality needed to process that document.

In conclusion and even though I am not a voting member right now, I
support the no-extension conformance class. I hope that it will
represent the way the the vast majority of documents are produced so
that we will have the greatest variety of consumers. The part I don't
like of the dual conformance clauses is that it even allows something
that has extensions to be called ODF. Defining this standard is, for
me, about interoperability and not about allowing mass divergence from
the standard through extensions. Otherwise, what point is there in
having the standard?


On Thu, Feb 26, 2009 at 18:56, Doug Mahugh <Doug.Mahugh@microsoft.com> wrote:
> I think Stephen answered your question, but I'd add that I never thought there was a compelling reason for two classes of conformance.  I could imagine a person wanting a particular extension, or not wanting it, but I can't imagine why a person would want to forego any and all extensions.  And even if a particular person wanted that (which I accept, since there are clearly person on the TC who seem to want that), I don't think we should add conformance classes to support all such possibilities.  Some users (working with the blind, for example, or working through extremely slow network connections) might prefer documents that don't have images in them, but I don't think we need a no-images conformance class either.
> Regards,
> Doug

To unsubscribe from this mail list, you must leave the OASIS TC that
generates this mail.  Follow this link to all your TCs in OASIS at:

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]